Ask your documents in natural language — answers based on their text
Ask a question in natural language about any document that has text (parsed or OCR): the answer is based strictly on that document, and the model says so when the answer is not in it. By default the model runs on your own server with Ollama (Llama, Mistral, Gemma… any Ollama model): your confidential data never leaves it. DeepSeek, OpenAI, Claude or Gemini can be chosen instead; the document text is then sent to that provider. The text given to the model is limited to a configurable size (12,000 characters by default, a few pages): raise it for long contracts. The questions and answers are kept on each document, with the user, the date and the model used.
Provided “as is”, without warranty. To the extent permitted by law, no liability is accepted for any damage arising from its use, including loss of data.
No refunds: the one-time payment is final and is not refunded, in whole or in part.
Exception: if the application does not work, or lacks a feature, compared to its description, a fix will be offered, or a refund if it cannot be fixed.